1 INFO-VAX	Sat, 26 Feb 2005	Volume 2005 : Issue 113       Contents: Backup /ALIAS question Re: Backup /ALIAS question Re: Backup /ALIAS question' Re: Checking for "file open for write". ' Re: Checking for "file open for write". ' Re: Checking for "file open for write". ' Re: Checking for "file open for write". ' Re: Checking for "file open for write". ' Re: Checking for "file open for write".  Re: FTP post processing  Re: FTP post processing  Re: Good case for VMS < Re: How to make BACKUP fail instead of making OPCOM request?< Re: How to make BACKUP fail instead of making OPCOM request?< Re: How to make BACKUP fail instead of making OPCOM request?
 HSZ70 Trouble  Re: HSZ70 Trouble  Re: HSZ70 Trouble  IA64 unraveling... Re: IA64 unraveling... Re: IA64 unraveling... Re: IA64 unraveling... Re: IA64 unraveling... Re: Install TCPIP 5.4  Re: Logfile name Re: Logfile name Re: Logfile name Re: Logfile name Re: Logfile name3 Re: London Stock Exchange slowing moving to Windows 3 RE: London Stock Exchange slowing moving to Windows 3 Re: London Stock Exchange slowing moving to Windows 3 RE: London Stock Exchange slowing moving to Windows 3 Re: London Stock Exchange slowing moving to Windows 3 Re: London Stock Exchange slowing moving to Windows 3 Re: London Stock Exchange slowing moving to Windows  Re: Mfg Date of an Alpha 2100 8 Re: OpenVMS Seminar in Toronto (2005-02-24) a few points8 Re: OpenVMS Seminar in Toronto (2005-02-24) a few points8 Re: OpenVMS Seminar in Toronto (2005-02-24) a few points Re: Ouch! a *MAJOR* bug in TPU.  Re: Ouch! a *MAJOR* bug in TPU.  Re: Ouch! a *MAJOR* bug in TPU.  Re: Ouch! a *MAJOR* bug in TPU.  Re: Point Secure Re: Sayonara Tukwilla  Re: Sayonara Tukwilla  Re: Sayonara Tukwilla  Re: Sayonara Tukwilla  Re: Sayonara Tukwilla 
 Re: ssl & osu 
 Re: ssl & osu 2 Re: WRQreflection 11: relocate settings directory?2 Re: WRQreflection 11: relocate settings directory?2 Re: WRQreflection 11: relocate settings directory?  F ----------------------------------------------------------------------  % Date: Fri, 25 Feb 2005 20:26:34 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com>  Subject: Backup /ALIAS question B Message-ID: <1109380457.8fbc6bef2a87186d3a7da440cb6bab7f@teranews>  H Help backup /alias is very confusing. What was the previous behaviour of backup which /alias preserves ?     & Would it be simpler to just state that  5 	/ALIAS backs up both the file entry and the contents G 	/NOALIAS backs up contents of primary file entry but only file entries  of aliases.   4 Or am I misunderstanding what /NOLIAS really means ?   ------------------------------    Date: 25 Feb 2005 18:50:41 -0800$ From: "AEF" <spamsink2001@yahoo.com># Subject: Re: Backup /ALIAS question C Message-ID: <1109386241.025012.241910@f14g2000cwb.googlegroups.com>    JF Mezei wrote: G > Help backup /alias is very confusing. What was the previous behaviour  of! > backup which /alias preserves ?  >  > ( > Would it be simpler to just state that > 7 > 	/ALIAS backs up both the file entry and the contents A > 	/NOALIAS backs up contents of primary file entry but only file  entries 
 > of aliases.  > 6 > Or am I misunderstanding what /NOLIAS really means ?    B Well, what does your HELP BACKUP /ALIAS say? It might vary amongst different versions of VMS.  F In VMS 6.2 I use /NOALIAS with image save operations. For some reason,C /IMAGE by itself saves all the primary files, but a also few of the B alias files (not all), at least according to the backup listings I made. So I just use   6     $ BACKUP/IMAGE/NOALIAS <disk-spec> <save-set-spec>  E plus any other quals I might want and I'm done with it. I can't speak  for other versions of VMS.  G The only reason I could see to use /ALIAS in an image save operation is B if you want to be able to selectively restore based on any a files aliases.  D The hp Web site docs recommend to specify /ALIAS only when restoring) very old save sets (VMS versions.LE.6.2).    ------------------------------  % Date: Fri, 25 Feb 2005 21:40:43 -0800  From: Z <Z@no.spam> # Subject: Re: Backup /ALIAS question + Message-ID: <n_TTd.10381$EF1.2050@fe02.lga>    JF Mezei wrote: J > Help backup /alias is very confusing. What was the previous behaviour of! > backup which /alias preserves ?  >  > ( > Would it be simpler to just state that > 7 > 	/ALIAS backs up both the file entry and the contents I > 	/NOALIAS backs up contents of primary file entry but only file entries 
 > of aliases.  > 6 > Or am I misunderstanding what /NOLIAS really means ?   Why are you asking this?  ; Don't you know that VMS commands like BACKUP are intuitive?    <stirs the pot>    ------------------------------  % Date: Fri, 25 Feb 2005 20:04:06 +0100 + From: Karsten Nyblad <nospam@nospam.nospam> 0 Subject: Re: Checking for "file open for write".- Message-ID: <cvnstd$2cpf$1@news.cybercity.dk>    prep@prep.synonet.com wrote:/ > Karsten Nyblad <nospam@nospam.nospam> writes: C >>The information on who has the file open is stored in the windows  >>control blocks (WCB).  >  > J > No, the WCB contains the virtual-physical mapping info. You are thinkingG > of the File Control Block. In a cluster, FCBs willonly exist on nodes C > that have the file open, so you can't `see' them without at least  > one open on the local node.   G No.  You should not try to discuss this subject with somebody who have  I actually written an application to get that information.  The is one WCB  E per channel but one FCB per file opened no matter how many times the  G file is opened.  The FCB does not have room for storing information on  D multiple channels.  Therefore that information is stored in the WCB. > B >> The file system does know who has the file open for write.  TheF >>locks are not normal locks.  After the lock on a file has been takenF >>out, the lock is converted into a lock, that is owned by the node ofE >>the cluster.  At least in the old days it was not documented how to B >>make such a convertion, but the flag is in starlet.  There is noE >>reason to scan the lock database your self.  SYS$GETLKI can get the B >>information you need.  And you can get the lock id from the fileA >>statistics block, which you get using the ACP QIO interface.  I D >>dont remember if you need to change to kernel mode before callingF >>SYS$GETLKI to get information on kernel mode locks.  Otherwise thereE >>is no need for kernel mode code.  You can write a utility that does C >>this with a call to SYS$GETLKI as the only code running in kernel  >>mode.  >  > D > IF you don't mind the corner case of a file opened `no lock', then4 > the locks on the file are the way to go for 99.9%. > H There is always a lock on an opened file.  Otherwise the information in > the file system could get out of date leading to loss of data.   ------------------------------  % Date: Sat, 26 Feb 2005 03:43:39 +0800  From: prep@prep.synonet.com 0 Subject: Re: Checking for "file open for write".- Message-ID: <87y8dc2zo4.fsf@prep.synonet.com>   - Karsten Nyblad <nospam@nospam.nospam> writes:   F > You have to understand that the VMS file system is based on the file> > system that existed before clusters.  Thus locks are used toG > synchronize access between nodes in a cluster, but synchronization of H > access on the single node is done by the file system without involving > locks.  C As of VMS 3.7, all file system sysncronisation has be done by locks 6 and the (D)LM. Both at the ACP/XQP level, and by RMS.      --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Sat, 26 Feb 2005 02:14:52 GMT   From: John Santos <john@egh.com>0 Subject: Re: Checking for "file open for write".+ Message-ID: <wYQTd.13041$QQ3.4329@trnddc02>    Hein wrote: 5 > "Paul Sture" <paul.sture@decus.ch> wrote in message ( > news:388av8F5dcr3qU1@individual.net... >  >>John Santos wrote: >  > I >>>I wonder if something like this could be hooked into the ACP's locking G >>>system or RMS or possibly by defining a default ACE on the receiving G >>>directory that would propagate to files created in the directory and C >>>trigger an alarm on file close that could be monitored by a user  >>>program?  >  > @ > Folks have done that. I'm sure there is stuff on the freeware.8 > Both with exploiting directory locks and audit events.G > But are we not over engineering this? Good simple solutions have been  > presented.9 > I have seen no indication where those solutions lacked.   ; (BTW, the issues here are more generic than just FTP.  They > also apply to NFS and, for that matter, to DECnet and RJ2780!)     Objections to the solutions:  I > -  Polling sounded reasonable (how long does the average tranfer take?)   # 1) 5GB file on a WAN takes a while! < 2) General philosophical objection to polling vs. interruptsA 3) What if something breaks and you are left with a partial file?   J > -  transferring with a temp name, and renaming when done is my preferred > technique.  G 1) This is my preferred method also, but the problem (as I've mentioned E before) is the client-side programmer forgetting why you are doing it G this way and breaking it by transferring it directly to the final name.   D If you have control over both the client and server, there shouldn'tA be a problem, but many times (for me) the issue has been that I'm A implementing one application for a customer and some other vendor F is implementing another application, and the two apps have to converse@ by exchanging files.  I remember why it works this way (and haveB documentation for the app that explains it), but can't control theB corporate memory of the other vendor.  Sometimes the customer willA replace the other side with a new product from yet another vendor ? that promises to be compatible, and the 1st we know about it is @ when the customer complains that our app is no longer processing the files correctly.  :-(   D 2) It is also a polling-based method.  (Picture Lisa and Bart in the8 back seat saying "Are we there yet?  Are we there yet?")  D 3) Without knowing the internals of the FTP server, is a file renameG an atomic operation?  Will your automated receive process always either F see a temp file (which it will ignore) or a complete, intact file withG the final name, and no other possibilities?  (For example, a brain-dead D FTP server might implement renames as copy old-name new-name, deleteF old-name, which would re-introduce the problems with polling, but with a smaller window.)  @ At least this method (and the marker file method) eliminate the G problem of something breaking and leaving a partially transmitted file.   N > - Those 'marker' files sound good also. ( A.dat ... realfile.dat ... Z.dat )  H 1) Also subject to human error if the sender forgets why the marker file: exists in the 1st place and sends them in the wrong order.   2) Also a polling method.   D This method is especially good when the sender is sending an unknownG number of files as a group.  (E.G. foo.1, foo.2, foo.3, ... foo.n.  The G application just waits for foo.flag to show up in the upload directory, G and then processes foo.*)  You only need one marker file for the whole   group.  J > - RMS knows how to make an RMS lock. It'll do it for you and not requireI > privs: Just try to open and you will learn the state. What's wrong with  > that?   F If your app opens the file "access read, allow none" in BASIC, (or FABA FAC=GET, SHR=NIL in RMS terms) and the FTP server creates it with G "access write, allow none (or read)" (FAB FAC=PUT, SHR=NIL or SHR=GET), " then this should work fine, but...  : 1) still has the "we only received a partial file" problem  H 2) It also relies on the FTP server behaving sanely, and not for exampleC creating the file (maybe pre-extending it), and then closing it and - re-opening it to do the actual data transfer.    > ? >>I like that idea, and RMS already provides the file attribute  >  > SUBMIT_ON_CLOSE. > ? >>How to set that attribute on an incoming file I am not  sure,  >  > N > Right, but that is a 'run time attribute', not a 'file permanent attribute'.N > So when FDL parses that command, it will flip a bit in the current fab and a$ > submit on close will be triggered.L > But the program having the file open has to set the bit. You can not use a > 'set file' or some such  >  > + >>but the following shows the general idea.  >  >  > reduced to bare-bones: >  > $ conv/fdl=tt: tt: tmp.com > file; submit yes > <Exit> > $show time > <exit> >  > 	 > Cheers,  > Hein.  >  >      --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Fri, 25 Feb 2005 22:13:49 -0500 ' From: Dave Froble <davef@tsoft-inc.com> 0 Subject: Re: Checking for "file open for write".0 Message-ID: <111vpqj3d3qc63b@corp.supernews.com>   John Santos wrote:  < > 1) still has the "we only received a partial file" problem  H For this reason I think the entire design of FTP is flawed.  In opening H an output file with the target name, without being assured that all the 6 contents of the file are in hand, FTP is a crap shoot.  D I've done data transfers many times in the past.  I've seen trading I partners with terrible plans, such as just shoot the data over the wire,  H the other side will pick it up, with no checking that the other side is  even there.   ) I prefer an handshaking protocol such as:   ) Sender> are you there and can I send data 2 Receiver> Yes, I am here and ready to receive data' Sender> <sends beginning of data token>  Sender> <sends data stream> ! Sender> <sends end of data token> B Sender> <sends information on data, byte count, record count, etc>: Receiver> <replies that information matches data received>   Or  G Receiver> <no, something is wrong> (and the transmission is treated an   if it never happened)   I If the transfer is Ok, then and only then does the receiver do something  5 with the data, such as create the target output file.   @ Ok, maybe I'm too picky, but I never have partial data problems.   Dave   ------------------------------  % Date: Sat, 26 Feb 2005 00:19:50 -0500 ' From: Dave Froble <davef@tsoft-inc.com> 0 Subject: Re: Checking for "file open for write".0 Message-ID: <112016rkp7i4e5f@corp.supernews.com>   prep@prep.synonet.com wrote:/ > Karsten Nyblad <nospam@nospam.nospam> writes:  >  > F >>You have to understand that the VMS file system is based on the file> >>system that existed before clusters.  Thus locks are used toG >>synchronize access between nodes in a cluster, but synchronization of H >>access on the single node is done by the file system without involving >>locks. >  > E > As of VMS 3.7, all file system sysncronisation has be done by locks 8 > and the (D)LM. Both at the ACP/XQP level, and by RMS.  >  > B Thanks for pointing this out.  I didn't want to get into a 1 on 1 $ shouting match over this issue.  :-)   ------------------------------  % Date: Sat, 26 Feb 2005 01:25:02 -0500 ( From: "Hein" <hein.nomail@hp.nomail.com>0 Subject: Re: Checking for "file open for write"., Message-ID: <422016c2$1@usenet01.boi.hp.com>  - "John Santos" <john@egh.com> wrote in message % news:wYQTd.13041$QQ3.4329@trnddc02... 
 > Hein wrote:  > Objections to the solutions: > K > > -  Polling sounded reasonable (how long does the average tranfer take?)  > % > 1) 5GB file on a WAN takes a while!   G But that's precisely why polling (as much as I dislike polling!), might  work. L If a transfer takes an hour, then polling every 30 seconds get you within 1%L accuracy of the real arrival time, and at a very manageable resource cost it would seem.   > > 2) General philosophical objection to polling vs. interrupts  & Agreed.. but if the price is right....G IMHO the interupt must come as a seperate, recognizable action from the  sender.   C > 3) What if something breaks and you are left with a partial file?   G How is an interupt, say from closing the file, going to be able to tell  whether it was partial or not?L It can tell cleaner, quicker that something happened, but not what happened.I Again, only the sender know, or needs to tell up fron what is coming such H that the receive can decide whether a valid transfer ha happened. At the8 very minimum a byte count, but probably a checksum also.  I > 1) This is my preferred method also, but the problem (as I've mentioned G > before) is the client-side programmer forgetting why you are doing it I > this way and breaking it by transferring it directly to the final name.   - Stick & Carrot. Make it their responsibility. ? You define how the receiver works and they'd better honor that. B > when the customer complains that our app is no longer processing > the files correctly.  :-(   K At which point you graceouisly point to the document/contract that explains K how your processing works, and show that it still works that way... if they  live by the rules.  F > 2) It is also a polling-based method.  (Picture Lisa and Bart in the: > back seat saying "Are we there yet?  Are we there yet?")  = File creates, and directory changes are easier to trigger on.   F > 3) Without knowing the internals of the FTP server, is a file rename > an atomic operation?  + VMS file renames are atomic (enough :-) :-)   L > > - RMS knows how to make an RMS lock. It'll do it for you and not requireK > > privs: Just try to open and you will learn the state. What's wrong with 	 > > that?  > H > If your app opens the file "access read, allow none" in BASIC, (or FABC > FAC=GET, SHR=NIL in RMS terms) and the FTP server creates it with I > "access write, allow none (or read)" (FAB FAC=PUT, SHR=NIL or SHR=GET), $ > then this should work fine, but...  * Not sure where you are going with that....  < > 1) still has the "we only received a partial file" problem   Right.  J > 2) It also relies on the FTP server behaving sanely, and not for exampleE > creating the file (maybe pre-extending it), and then closing it and / > re-opening it to do the actual data transfer.   H Yes, but any RMS/XQP/xxx lock method discussed before suffers from that.H I'm just saying that if you consider using the fle locking, then just go RMS.  D Odd, how this 30+ year old problem has not been solved properly huh?K VMS had/has 'SPOOL' nee FTSV (1985-ish in Valbonne :-) to deal with most of H these issues, including a 'trigger when done, post processing function'.   Cheers,  Hein.    ------------------------------  + Date: Fri, 25 Feb 2005 16:43:30 -0600 (CST) * From: sms@antinode.org (Steven M. Schweda)  Subject: Re: FTP post processing2 Message-ID: <05022516433051_27800279@antinode.org>  4 From: karcher@thuria.waisman.wisc.edu (Carl Karcher)    > [...] I'd rather see resourcesC > expended on adding a feature to the ftp server that would make it G > function well with GUI ftp clients when a directory name contains a ^ F > character. A unix emulation mode perhaps. I've yet to find a GUI ftpB > client that will transfer (or even display) files in a directoryG > with ^ in the name. Not that I care much about GUI ftp clients but my  > users do.   A    As I was starting to become satisfied with ODS5 support in the H Info-ZIP [Un]Zip programs, I just started looking at ODS5 v. Wget, whichB requires dealing with the escaped name characters in the DIRECTORY$ listing (FTP "LIST" command output).  >    From this point of view, the big problem is not the escapedE characters in ODS5 extended file names, it's the unwillingness of the D FTP server (TCPIP V5.4 - ECO 4) to _accept_ escaped characters.  ForE example, it may show a directory as "a^[b^]c.DIR;1", but just try and E use that name as-is (shown here using the matching TCPIP FTP client):    FTP> cd a^[b^]c / 550-Failed to set default directory to a^[b^]c.  550 invalid directory   E If the escaped characters are de-escaped (re-imprisoned?), it's fine:   
 FTP> cd a[b]c  250-CWD command successful. 1 250 New default directory is LDA1:[test2.a^[b^]c]   7    Interestingly, a directory spec with brackets is ok:    FTP> cd [.a^[b^]c] 250-CWD command successful. 1 250 New default directory is LDA1:[test2.a^[b^]c]   G    Similar results are obtained for non-directory files.  Naturally, if B a file name contains an escaped slash (which, I'll admit, is a badG idea), you're absolutely doomed.  For a plain file, expect results like  these:   FTP> get a^.b.c;1  200 PORT command successful.3 550-Failed to open LDA1:[test]a^^^.b.c;1 for input.  550 file not found   FTP> get a.b.c 200 PORT command successful.N 150 Opening data connection for LDA1:[test]a^.b.c; (10.0.0.9,55879) (12 bytes) 226 Transfer complete.+ local: ALP$DKA0:[SMS]A_B.C;1  remote: a.b.c 8 12 bytes received in 00:00:00.02 seconds (0.56 Kbytes/s)  B    Nice renaming for the ODS2 destination, at least.  (Some of theC trouble shown here may be blamed on the client, but my work on Wget ? suggests that the server is the main source of the difficulty.)   B    A program which will not tolerate as input its own output lacksH something, possibly self-respect.  It also demands special handling fromH any other program which might wish to interact with it.  It ill behoovesF VMS to make life difficult for users of other systems which might wish to interact with it.  G    Adventurous Wget users with time to expend might wish to investigate D the latest set of patches at "http://antinode.org/dec/sw/wget.html",A which just _may_ be able to do FTP with the type of FTP server in 	 question.   H ------------------------------------------------------------------------  4    Steven M. Schweda               (+1) 651-699-98183    382 South Warwick Street        sms@antinode-org     Saint Paul  MN  55105-2547    ------------------------------  % Date: Fri, 25 Feb 2005 21:06:12 -0600 2 From: David J Dachtera <djesys.nospam@comcast.net>  Subject: Re: FTP post processing+ Message-ID: <421FE7A4.E4601FDA@comcast.net>    Carl Karcher wrote:  > J > ->> This comes up so often, perhaps it is time to tackle this and make a( > ->> formal request to VMS management ? > H > While this would be a useful feature to some, I'd rather see resourcesC > expended on adding a feature to the ftp server that would make it G > function well with GUI ftp clients when a directory name contains a ^ F > character. A unix emulation mode perhaps. I've yet to find a GUI ftpB > client that will transfer (or even display) files in a directoryG > with ^ in the name. Not that I care much about GUI ftp clients but my  > users do.   G Even more reason to stay with RFTP V6.10 (THE usable one, not V6.20 and  later).    --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Fri, 25 Feb 2005 23:00:10 -0500 ' From: Dave Froble <davef@tsoft-inc.com>  Subject: Re: Good case for VMS0 Message-ID: <111vshgkugu095a@corp.supernews.com>   DeanW wrote:E > Anyone else notice that the BOINC version of Seti@Home is down? I'm A > not going to claim that VMS would have prevented all the issues F > they're having right now- but it might have helped. I leave it as anE > exercise to the reader to determine the cost:benefit ratio a 2-node ? > cluster that would have prevented this whole mess might have.   E Forget the cluster.  Just consider a file system that thinks data is  6 important enough to get it to disk, not just in cache.  I There may be some applications where security of data isn't as important   as speed, but not many.    Dave   ------------------------------  % Date: Fri, 25 Feb 2005 20:21:17 -0600 2 From: David J Dachtera <djesys.nospam@comcast.net>E Subject: Re: How to make BACKUP fail instead of making OPCOM request? + Message-ID: <421FDD1D.81004D94@comcast.net>    John Vottero wrote:  > A > "David J Dachtera" <djesys.nospam@comcast.net> wrote in message ' > news:421E965D.F8E1C305@comcast.net...  > > Bob Koehler wrote: > >>J > >> In article <1109201710.f768b09e46e168f6fb6433c34a6a9c8e@teranews>, JF1 > >> Mezei <jfmezei.spamnot@teksavvy.com> writes: J > >> > Different slant: would it be possible for your batch job to spawn aJ > >> > subprocess which VMS would view as "interactive" at which point the, > >> > /NOASSIST might be handled properly ? > >>K > >>    Getting the subprocess to be "interactive" is a non-trivial effort.  > > * > > Maybe play some games with FT devices? > > K > > We've needed for a long time to have a version of (ALPHA_)LOGGER.C that 6 > > would work from a file as well as from a terminal. > >  > L > Are you looking for a way to run an interactive application with the input > coming from a file?   E Actually, what is needed is a way to SET HOST and other commands with H SYS$INPUT pointing to a file rather than the terminal. (HSZTERM$SCSIPAD,F aka "SET HOST/SCSI" already does this, but is unique in that respect).C Likewise, we need the equivalent of (ALPHA_)LOGGER to have the same  capability.    --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Sat, 26 Feb 2005 02:47:06 +0000 - From: David B Sneddon <dbsneddon@bigpond.com> E Subject: Re: How to make BACKUP fail instead of making OPCOM request? * Message-ID: <421FE32A.4090207@bigpond.com>  & David J Dachtera was overheard to say:G > Actually, what is needed is a way to SET HOST and other commands with J > SYS$INPUT pointing to a file rather than the terminal. (HSZTERM$SCSIPAD,H > aka "SET HOST/SCSI" already does this, but is unique in that respect).E > Likewise, we need the equivalent of (ALPHA_)LOGGER to have the same 
 > capability.  >    Kermit script?   Regards, Dave.  --  D David B Sneddon (dbs)  VMS Systems Programmer  dbsneddon@bigpond.comD Sneddo's quick guide ...     http://www.users.bigpond.com/dbsneddon/D DBS freeware     http://www.users.bigpond.com/dbsneddon/software.htm   ------------------------------  % Date: Fri, 25 Feb 2005 21:20:28 -0600 2 From: David J Dachtera <djesys.nospam@comcast.net>E Subject: Re: How to make BACKUP fail instead of making OPCOM request? + Message-ID: <421FEAFC.510F64ED@comcast.net>    David B Sneddon wrote: > ( > David J Dachtera was overheard to say:I > > Actually, what is needed is a way to SET HOST and other commands with L > > SYS$INPUT pointing to a file rather than the terminal. (HSZTERM$SCSIPAD,J > > aka "SET HOST/SCSI" already does this, but is unique in that respect).G > > Likewise, we need the equivalent of (ALPHA_)LOGGER to have the same  > > capability.  > >  >  > Kermit script?   No. I need this:  $ $ DEFINE/USER SYS$OUTPUT output_file! $ DEFINE/USER SYS$INPUT some_file 
 $ MCR LOGGER    0 ...where "some_file" might contain, for example,  * SET HOST/DUP/SERV=MSCP$DUP/TASK=CLI HSJ004 SHOW THIS FULL SHOW OTHER FULL 
 SHOW UNITS SHOW UNITS FULL  SHOW STORAGESETS SHOW STORAGESETS FULL  SHOW DEVICES SHOW DEVICES FULL  EXIT  B Yes, I know about HSDSA-SCRIPT. You'll find an archive of it at my freeware site.   --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------    Date: 25 Feb 2005 14:17:27 -0800" From: "Jim  M" <jckstrw@gmail.com> Subject: HSZ70 TroubleC Message-ID: <1109369847.227619.105870@z14g2000cwz.googlegroups.com>    Hi  D I've recently inherited an AlphaServer 4100 with an HSZ70 controllerF and several arrays.  The system was working fine without any issues upC until the system was moved from one server room to another.  It was ? moved at the beginning of this year and just yesterday we tried  bringing the system up.   @ I had some issues trying to bring up the system. Because the hszC controller was off for so long the battery cache was depleted and I F needed to clear the errors. I did that and when I try and bring up theD box (with just 'boot' at the >>> prompt), I get a lot of errors likeE "dkc0.0.0.3.1 media not present or disabled via the RUN/STOP switch".   G I'm not sure what's wrong (it's been a really long time since I did VMS F work). The hsz controller isn't showing any OP error codes. I've triedC to do some troubleshooting on the controller, but I'm stuck. Here's E what I've got--I've run through several of the self-tests and the one 6 that gives me pause is the "selftest this_controller":    HSZ70 > selftest this_controllerE Error 4160: Unable to rundown the following units on this controller:  D0 -- Cannot rundown unit  D100 -- Cannot rundown unit  D200 -- Cannot rundown unit  D201 -- Cannot rundown unit  D302 -- Cannot rundown unit  D102 -- Cannot rundown unit  D301 -- Cannot rundown unit  D104 -- Cannot rundown unit  D304 -- Cannot rundown unit  D303 -- Cannot rundown unit  D300 -- Cannot rundown unit   E Here's some more (truncated a bit, but all the devices show basically  the same details):   HSZ70 > show disks full E Name          Type                      Port Targ  Lun        Used by N ------------------------------------------------------------------------------  @ DISK10000     disk                         1    0    0        M1(           DEC      RZ1CB-CA (C) DEC LYJ0         Switches:            NOTRANSPORTABLE @           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ negotiated)          Size: 8378028 blocksA DISK10100     disk                         1    1    0        M11 (           COMPAQ   BD018635C4       B017         Switches:            NOTRANSPORTABLE @           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ negotiated)          Size: 35556389 blocks @ DISK10200     disk                         1    2    0        M3(           DEC      RZ1DB-CA (C) DEC LYJ0         Switches:            NOTRANSPORTABLE @           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ negotiated)          Size: 17769177 blocks @ DISK10300     disk                         1    3    0        M4(           DEC      RZ1DB-CA (C) DEC LYJ0         Switches:            NOTRANSPORTABLE @           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ negotiated)          Size: 17769177 blocks @ DISK20000     disk                         2    0    0        M1(           DEC      RZ1CB-CA (C) DEC LYJ0         Switches:            NOTRANSPORTABLE @           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ negotiated)          Size: 8378028 blocks  ' Again, here's some more truncated info:    HSZ70 > show unit full1     LUN                                      Uses > --------------------------------------------------------------  /   D0                                         M1          Switches: B           RUN                    NOWRITE_PROTECT        READ_CACHE             WRITEBACK_CACHE +           MAXIMUM_CACHED_TRANSFER_SIZE = 32            ACCESS_ID = ALL          State:           INOPERATIVE            Unit has lost data*           PREFERRED_PATH = THIS_CONTROLLER%           WRITE_PROTECT - DATA SAFETY          Size: NOT YET KNOWN /   D100                                       M2          Switches: B           RUN                    NOWRITE_PROTECT        READ_CACHE             WRITEBACK_CACHE +           MAXIMUM_CACHED_TRANSFER_SIZE = 32            ACCESS_ID = ALL          State:           INOPERATIVE            Unit has lost data*           PREFERRED_PATH = THIS_CONTROLLER%           WRITE_PROTECT - DATA SAFETY          Size: NOT YET KNOWN /   D102                                       M6          Switches: B           RUN                    NOWRITE_PROTECT        READ_CACHE             WRITEBACK_CACHE +           MAXIMUM_CACHED_TRANSFER_SIZE = 32            ACCESS_ID = ALL          State:           INOPERATIVE	           Unit has lost data*           PREFERRED_PATH = THIS_CONTROLLER%           WRITE_PROTECT - DATA SAFETYu         Size: NOT YET KNOWNe/   D104                                       M8o         Switches:cB           RUN                    NOWRITE_PROTECT        READ_CACHE             WRITEBACK_CACHEe+           MAXIMUM_CACHED_TRANSFER_SIZE = 32e           ACCESS_ID = ALLt         State:           INOPERATIVE            Unit has lost data*           PREFERRED_PATH = THIS_CONTROLLER%           WRITE_PROTECT - DATA SAFETY          Size: NOT YET KNOWNs  E What else can I do?  If you need more information please let me know.r   Thanks.    Jimr   ------------------------------  # Date: Sat, 26 Feb 2005 00:02:00 GMTe8 From: "Schroeder, AJ" <aaron.schroeder@tmscomputers.com> Subject: Re: HSZ70 Trouble4 Message-ID: <Y%OTd.133$Cg6.65@twister.rdc-kc.rr.com>   Jim M wrote: > Hi > F > I've recently inherited an AlphaServer 4100 with an HSZ70 controllerH > and several arrays.  The system was working fine without any issues upE > until the system was moved from one server room to another.  It wasiA > moved at the beginning of this year and just yesterday we tried  > bringing the system up.  > B > I had some issues trying to bring up the system. Because the hszE > controller was off for so long the battery cache was depleted and IoH > needed to clear the errors. I did that and when I try and bring up theF > box (with just 'boot' at the >>> prompt), I get a lot of errors likeG > "dkc0.0.0.3.1 media not present or disabled via the RUN/STOP switch".h > I > I'm not sure what's wrong (it's been a really long time since I did VMSeH > work). The hsz controller isn't showing any OP error codes. I've triedE > to do some troubleshooting on the controller, but I'm stuck. Here'ssG > what I've got--I've run through several of the self-tests and the one 8 > that gives me pause is the "selftest this_controller": > " > HSZ70 > selftest this_controllerG > Error 4160: Unable to rundown the following units on this controller:, > D0 -- Cannot rundown unit  > D100 -- Cannot rundown unit@ > D200 -- Cannot rundown unit  > D201 -- Cannot rundown unit1 > D302 -- Cannot rundown unit4 > D102 -- Cannot rundown unitc > D301 -- Cannot rundown unith > D104 -- Cannot rundown unitf > D304 -- Cannot rundown unit  > D303 -- Cannot rundown unitj > D300 -- Cannot rundown unit  > G > Here's some more (truncated a bit, but all the devices show basicallya > the same details): >  > HSZ70 > show disks full G > Name          Type                      Port Targ  Lun        Used by-P > ------------------------------------------------------------------------------ > B > DISK10000     disk                         1    0    0        M1* >           DEC      RZ1CB-CA (C) DEC LYJ0 >         Switches:e >           NOTRANSPORTABLEuB >           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ
 > negotiated)  >         Size: 8378028 blocksC > DISK10100     disk                         1    1    0        M11e* >           COMPAQ   BD018635C4       B017 >         Switches:y >           NOTRANSPORTABLEsB >           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ
 > negotiated)  >         Size: 35556389 blocksiB > DISK10200     disk                         1    2    0        M3* >           DEC      RZ1DB-CA (C) DEC LYJ0 >         Switches:e >           NOTRANSPORTABLEsB >           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ
 > negotiated)i >         Size: 17769177 blocks B > DISK10300     disk                         1    3    0        M4* >           DEC      RZ1DB-CA (C) DEC LYJ0 >         Switches:  >           NOTRANSPORTABLE B >           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ
 > negotiated)o >         Size: 17769177 blocksaB > DISK20000     disk                         2    0    0        M1* >           DEC      RZ1CB-CA (C) DEC LYJ0 >         Switches:s >           NOTRANSPORTABLE-B >           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ
 > negotiated)j >         Size: 8378028 blocks > ) > Again, here's some more truncated info:  >  > HSZ70 > show unit full3 >     LUN                                      Usese@ > -------------------------------------------------------------- > 1 >   D0                                         M1b >         Switches:dD >           RUN                    NOWRITE_PROTECT        READ_CACHE >  >           WRITEBACK_CACHEs- >           MAXIMUM_CACHED_TRANSFER_SIZE = 32  >           ACCESS_ID = ALL  >         State: >           INOPERATIVEm >           Unit has lost data, >           PREFERRED_PATH = THIS_CONTROLLER' >           WRITE_PROTECT - DATA SAFETY0 >         Size: NOT YET KNOWN<1 >   D100                                       M2f >         Switches: D >           RUN                    NOWRITE_PROTECT        READ_CACHE >  >           WRITEBACK_CACHEm- >           MAXIMUM_CACHED_TRANSFER_SIZE = 32w >           ACCESS_ID = ALLd >         State: >           INOPERATIVE  >           Unit has lost data, >           PREFERRED_PATH = THIS_CONTROLLER' >           WRITE_PROTECT - DATA SAFETYl >         Size: NOT YET KNOWNo1 >   D102                                       M6  >         Switches: D >           RUN                    NOWRITE_PROTECT        READ_CACHE >  >           WRITEBACK_CACHEa- >           MAXIMUM_CACHED_TRANSFER_SIZE = 32a >           ACCESS_ID = ALLB >         State: >           INOPERATIVE  >           Unit has lost data, >           PREFERRED_PATH = THIS_CONTROLLER' >           WRITE_PROTECT - DATA SAFETYn >         Size: NOT YET KNOWNn1 >   D104                                       M8  >         Switches: D >           RUN                    NOWRITE_PROTECT        READ_CACHE >  >           WRITEBACK_CACHEe- >           MAXIMUM_CACHED_TRANSFER_SIZE = 32y >           ACCESS_ID = ALL  >         State: >           INOPERATIVEt >           Unit has lost data, >           PREFERRED_PATH = THIS_CONTROLLER' >           WRITE_PROTECT - DATA SAFETY  >         Size: NOT YET KNOWN  > G > What else can I do?  If you need more information please let me know.i > 	 > Thanks.  >  > Jimn >   H I haven't administered these disk subsystems for about 3 years now, but / I remember the CLI like the back of my hand. :)o  A As you stated, this system was shutdown improperly and the cache eC batteries were not fully discharged. That means that each unit has $I unflushed data, meaning that the HSZ controllers had cached data waiting tB to be purged to disk, and since the controllers were not shutdown H properly, the HSZ subsystem has placed a write lock on all units, hence D the WRITE_PROTECT, Data Safety, and most importantly "unit has lost 
 data" output.e  L When a HSZ or HSG controller tells you that, you need to issue this command:  . CLEAR <UNIT> LOST_DATA DESTROY_UNFLUSHED_CACHE  @ and repeat for each unit that is reporting as offline or in the H WRITE_PROTECT mode. Then reboot your server and things should be hunky-  dory.   H There is also a NODESTROY_UNFLUSHED_DATA, but that is pretty useless at - this point since your batteries are way dead.a  H Note: In the future, to properly shutdown a ESA, you need to (depending E on your config) tell each controller to shutdown by doing a SHUTDOWN tC OTHER and then SHUTDOWN THIS. Then you need to turn off your cache nG batteries by holding the black button on the battery unit itself until RG the LED stops flashing. Then the batteries are discharged and ready to dE go. Simply turning these towers off results in the condition you are  @ experiencing. Of course now you might need to get new batteries.   HTH,   AJ Schroeder   ------------------------------  % Date: Fri, 25 Feb 2005 17:03:25 -0700c+ From: John Nebel <john.nebel_vms@csdco.com>i Subject: Re: HSZ70 Trouble& Message-ID: <421FBCCD.70402@csdco.com>   Jim,  A Use the clear_errors unit# lost_data command.  If the disks were lH dismounted properly and the array were left powered on for more than theK cache_flush_timer (default 10 seconds) thereafter, the disks are likely OK.F   John   Jim M wrote:   > Hi > F > I've recently inherited an AlphaServer 4100 with an HSZ70 controllerH > and several arrays.  The system was working fine without any issues upE > until the system was moved from one server room to another.  It wassA > moved at the beginning of this year and just yesterday we triedh > bringing the system up.u > B > I had some issues trying to bring up the system. Because the hszE > controller was off for so long the battery cache was depleted and IeH > needed to clear the errors. I did that and when I try and bring up theF > box (with just 'boot' at the >>> prompt), I get a lot of errors likeG > "dkc0.0.0.3.1 media not present or disabled via the RUN/STOP switch".  > I > I'm not sure what's wrong (it's been a really long time since I did VMStH > work). The hsz controller isn't showing any OP error codes. I've triedE > to do some troubleshooting on the controller, but I'm stuck. Here'slG > what I've got--I've run through several of the self-tests and the oneb8 > that gives me pause is the "selftest this_controller": > " > HSZ70 > selftest this_controllerG > Error 4160: Unable to rundown the following units on this controller:a > D0 -- Cannot rundown unit  > D100 -- Cannot rundown unitn > D200 -- Cannot rundown unity > D201 -- Cannot rundown unitd > D302 -- Cannot rundown unit  > D102 -- Cannot rundown unit  > D301 -- Cannot rundown unith > D104 -- Cannot rundown unitl > D304 -- Cannot rundown unitm > D303 -- Cannot rundown unitt > D300 -- Cannot rundown uniti > G > Here's some more (truncated a bit, but all the devices show basicallyt > the same details): >  > HSZ70 > show disks full G > Name          Type                      Port Targ  Lun        Used by P > ------------------------------------------------------------------------------ > B > DISK10000     disk                         1    0    0        M1* >           DEC      RZ1CB-CA (C) DEC LYJ0 >         Switches:h >           NOTRANSPORTABLEhB >           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ
 > negotiated). >         Size: 8378028 blocksC > DISK10100     disk                         1    1    0        M11 * >           COMPAQ   BD018635C4       B017 >         Switches:r >           NOTRANSPORTABLEfB >           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ
 > negotiated)  >         Size: 35556389 blocksiB > DISK10200     disk                         1    2    0        M3* >           DEC      RZ1DB-CA (C) DEC LYJ0 >         Switches:p >           NOTRANSPORTABLEmB >           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ
 > negotiated)u >         Size: 17769177 blockslB > DISK10300     disk                         1    3    0        M4* >           DEC      RZ1DB-CA (C) DEC LYJ0 >         Switches:  >           NOTRANSPORTABLEsB >           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ
 > negotiated)r >         Size: 17769177 blocksmB > DISK20000     disk                         2    0    0        M1* >           DEC      RZ1CB-CA (C) DEC LYJ0 >         Switches:d >           NOTRANSPORTABLEnB >           TRANSFER_RATE_REQUESTED = 20MHZ (synchronous 20.00 MHZ
 > negotiated)  >         Size: 8378028 blocks > ) > Again, here's some more truncated info:h >  > HSZ70 > show unit full3 >     LUN                                      UsesM@ > -------------------------------------------------------------- > 1 >   D0                                         M1a >         Switches: D >           RUN                    NOWRITE_PROTECT        READ_CACHE >  >           WRITEBACK_CACHEi- >           MAXIMUM_CACHED_TRANSFER_SIZE = 32  >           ACCESS_ID = ALLo >         State: >           INOPERATIVE  >           Unit has lost data, >           PREFERRED_PATH = THIS_CONTROLLER' >           WRITE_PROTECT - DATA SAFETY  >         Size: NOT YET KNOWNl1 >   D100                                       M2e >         Switches:hD >           RUN                    NOWRITE_PROTECT        READ_CACHE >  >           WRITEBACK_CACHEl- >           MAXIMUM_CACHED_TRANSFER_SIZE = 32T >           ACCESS_ID = ALLt >         State: >           INOPERATIVE  >           Unit has lost data, >           PREFERRED_PATH = THIS_CONTROLLER' >           WRITE_PROTECT - DATA SAFETYa >         Size: NOT YET KNOWNn1 >   D102                                       M6r >         Switches:rD >           RUN                    NOWRITE_PROTECT        READ_CACHE >  >           WRITEBACK_CACHE>- >           MAXIMUM_CACHED_TRANSFER_SIZE = 32  >           ACCESS_ID = ALL  >         State: >           INOPERATIVEu >           Unit has lost data, >           PREFERRED_PATH = THIS_CONTROLLER' >           WRITE_PROTECT - DATA SAFETYh >         Size: NOT YET KNOWNt1 >   D104                                       M8, >         Switches: D >           RUN                    NOWRITE_PROTECT        READ_CACHE >  >           WRITEBACK_CACHEs- >           MAXIMUM_CACHED_TRANSFER_SIZE = 32  >           ACCESS_ID = ALLi >         State: >           INOPERATIVE" >           Unit has lost data, >           PREFERRED_PATH = THIS_CONTROLLER' >           WRITE_PROTECT - DATA SAFETYm >         Size: NOT YET KNOWNl > G > What else can I do?  If you need more information please let me know.a > 	 > Thanks.n >  > Jimt   ------------------------------  % Date: Fri, 25 Feb 2005 21:04:25 -0500o- From: JF Mezei <jfmezei.spamnot@teksavvy.com>l Subject: IA64 unraveling...gB Message-ID: <1109382724.4273615e1a8dca69d83d56c8c5b058ce@teranews>  c > http://news.com.com/Sources+IBM+ditching+Itanium+altogether/2100-1006_3-5590780.html?tag=nefd.tope  F This one announces that IBM is not likely to produce new IA64 servers. (Not official yet).n  8 But there is a more interesting set of paragraphs in it:   ##E Ceasing chipset development doesn't preclude IBM from selling Itaniuma> servers. But it would need to use a server design from anotherG source--either a comparatively neutral supplier such as Intel or an IBM E rival that offers Itanium chipsets: HP, Fujitsu, SGI, NEC, Hitachi ort Unisys.   G But Intel isn't a likely Itanium chipset source. It had plans for a newvD chipset code-named Bayshore for Montecito-based Itanium servers, but0 canceled those plans in 2004, the company said.   F Montecito will still work with Intel's existing 8870 chipset, but that@ model was introduced in 2002 and lacks support for newer, fasterG communications technologies such as DDR2 (double data rate 2) memory or > PCI Express input-output. Intel will not be updating the 8870, spokeswoman Erica Fields said. s ##    D IBM dropping support for IA64 is no big surprise. But Intel droppingD development of support chips to make better use of Montecito ??? NotH exactly great news. Has Intel announced other new chipsets for montecito  to replace the one they canned ?  C So far, nothing has been done to kill off the idea that Intel is on & course to wind down this IA64 project.   ------------------------------  % Date: Fri, 25 Feb 2005 21:57:59 -0500i( From: Bill Todd <billtodd@metrocast.net> Subject: Re: IA64 unraveling...5= Message-ID: <spWdnY2-YJgleILfRVn-rA@metrocastcablevision.com>R   JF Mezei wrote:l   ...f  F > IBM dropping support for IA64 is no big surprise. But Intel droppingF > development of support chips to make better use of Montecito ??? Not > exactly great news.e  G Not exactly new news, either:  Intel canceled those chipsets late last oE summer, IIRC (around the time when it started to become obvious that eC Pentium-M cores rather than current P4/Prescott cores were the x86 ) future at Intel).f  6   Has Intel announced other new chipsets for montecito" > to replace the one they canned ?  D My impression is that they haven't:  they seem to be depending upon E others for anything better than the existing chipsets for Montecito, gH despite still planning to offer Montecito in a new (faster bus) variety E that the existing chipsets can't effectively make use of (so I guess gE they expect *someone* to have a new chipset out there to use it - or aH perhaps they have something of their own coming in the 4-processor size  range but nothing larger).   > E > So far, nothing has been done to kill off the idea that Intel is one( > course to wind down this IA64 project.  G They certainly aren't talking about anything new on the horizon beyond pI Tukwila (and Tukwila is now a lot less 'new' than it once was).  If they  G aren't actually planning to wind Itanic down, they at least seem to be  F taking much more of a wait-and-see attitude rather than forging ahead B anywhere nearly as aggressively as they have for the past 5 years.   - bill   ------------------------------  % Date: Fri, 25 Feb 2005 22:12:43 -0500e( From: Bill Todd <billtodd@metrocast.net> Subject: Re: IA64 unraveling...o= Message-ID: <i_GdnZ7i2eaxdILfRVn-sA@metrocastcablevision.com>    Bill Todd wrote: > JF Mezei wrote:  >  > ...t > G >> IBM dropping support for IA64 is no big surprise. But Intel droppingoG >> development of support chips to make better use of Montecito ??? Not- >> exactly great news. >  > I > Not exactly new news, either:  Intel canceled those chipsets late last  G > summer, IIRC (around the time when it started to become obvious that tE > Pentium-M cores rather than current P4/Prescott cores were the x86 o > future at Intel).   K Should have Googled it up first - turns out to have been early last summer:l  ) http://www.theinquirer.net/?article=16868    - bill   ------------------------------  % Date: Fri, 25 Feb 2005 23:06:18 -0500 ' From: Dave Froble <davef@tsoft-inc.com>r Subject: Re: IA64 unraveling...B0 Message-ID: <111vssve0b3hd81@corp.supernews.com>   JF Mezei wrote: c >>http://news.com.com/Sources+IBM+ditching+Itanium+altogether/2100-1006_3-5590780.html?tag=nefd.top  >  > H > This one announces that IBM is not likely to produce new IA64 servers. > (Not official yet).i > : > But there is a more interesting set of paragraphs in it: >  > ##G > Ceasing chipset development doesn't preclude IBM from selling Itaniumi@ > servers. But it would need to use a server design from anotherI > source--either a comparatively neutral supplier such as Intel or an IBMrG > rival that offers Itanium chipsets: HP, Fujitsu, SGI, NEC, Hitachi or 
 > Unisys.  > I > But Intel isn't a likely Itanium chipset source. It had plans for a new F > chipset code-named Bayshore for Montecito-based Itanium servers, but2 > canceled those plans in 2004, the company said.  > H > Montecito will still work with Intel's existing 8870 chipset, but thatB > model was introduced in 2002 and lacks support for newer, fasterI > communications technologies such as DDR2 (double data rate 2) memory or0@ > PCI Express input-output. Intel will not be updating the 8870,! > spokeswoman Erica Fields said. . > ## >  > F > IBM dropping support for IA64 is no big surprise. But Intel droppingF > development of support chips to make better use of Montecito ??? NotJ > exactly great news. Has Intel announced other new chipsets for montecito" > to replace the one they canned ? > E > So far, nothing has been done to kill off the idea that Intel is onc( > course to wind down this IA64 project.  F Ask Rob, he knows everything.  I'm sure he'll be able to spin this to   something that sounds good.  :-)   ------------------------------    Date: 25 Feb 2005 22:39:15 -0800 From: icerq4a@spray.se Subject: Re: IA64 unraveling...PC Message-ID: <1109399955.692025.173500@f14g2000cwb.googlegroups.com>e   JF Mezei wrote:d >h > But Intel droppingF > development of support chips to make better use of Montecito ??? Not@ > exactly great news. Has Intel announced other new chipsets for	 montecitoy" > to replace the one they canned ?  C This is old news, I think early 2004. Intel has not announced a new)E chipset, but atleast HP and SGI have new chipsets in the works. Intel @ has more or less said that they focus on the common Xeon/Itanium chipset instead.  9 > So far, nothing has been done to kill off the idea that 4 > Intel is on course to wind down this IA64 project.  : Some things have been done. The announcement of the commonE platform/chipsets for Xeon and Itanium in the future. This is perhapst) something Intel should have done earlier.r  G Think about how many Proliants HP ship, when they in the future will beoD able to take IA64 chips... Prices will go down, and there is atleast much more potential for volume.e   ------------------------------  # Date: Sat, 26 Feb 2005 01:08:36 GMTc  From: John Santos <john@egh.com> Subject: Re: Install TCPIP 5.4* Message-ID: <o_PTd.17985$uc.7028@trnddc09>   JF Mezei wrote:. > Tom Linden wrote:n > B >>5.1 came in my 7.3 kit.  Do you know if SSH2 is client/server is >>available for 5.3? >  > - > Not on VAX,. I think this is a 5.4 feature.n  A FYI the VAX SPL Q1CY2005, which I received last week, has "Compaq B TCP/IP Services for OpenVMS" 5.3 on it.  (They haven't changed theA name yet :-)  It's not marked as "NEW" in the master index, and I A think it has been out for a while.  So as of last week, at least,  no 5.4 yet.n  D As Dan O'Reilly said in another followup, TCPWare (and Multinet too,E I think) has had SSH2 for VAX for a while.  I've used it occasionallya without any obvious problems.M  ? The Process FTP site had patch kit SSH_V562P050 on it, but thathC has disappeared and ...P040 is there now.  I would guess this meansnC they found a problem with P050 and withdrew it, but I haven't heardtA anything official or unofficial.  Also, P050 still seems to be ontG the web site.  (http://www.process.com/techsupport/patches.html#tcpware-9 vs. ftp.process.com ANONYMOUS_ROOT:[000000.SUPPORT.56_2])S  F Process also has an SSH product that runs on VAX on top of UCX/TCP/IP, but I have never used it.t   -- M John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------    Date: 25 Feb 2005 11:11:50 -0800$ From: "AEF" <spamsink2001@yahoo.com> Subject: Re: Logfile name4C Message-ID: <1109358709.937725.240860@o13g2000cwo.googlegroups.com>    Ken Fairfield wrote: > AEF wrote: > > David Gray wrote:  > >l > >>Hi all,M > >>C > >>From within a batch job how do I find out what the value of theh /LOG! > >>qualifier was at submit time?f > >> > >>e.g. > >>@ > >>$ submit/log=gen_logs:a_log_file_with_a_long_name.log	 X.COM > >>F > >>When X.COM is running I want to put the value specifed on the /LOG > >>into a symbol. > >> > >>$ show symbol LogFile = > >>     LogFile = "gen_logs:a_log_file_with_a_long_name.log"o > >>A > >>I was fairly sure I could do this with a lexical function butr cannot > >>find/remember which one. > >>
 > >>Cheers
 > >>	Dave. > >  > >- > >a
 > > Try this:  > >J > > $    SET VERIFYw > > $    LOG_FILE =i7 F$GETQUI("DISPLAY_JOB","LOG_SPECIFICATION",,"THIS_JOB")o > > $    SHOW SYMBOL LOG_FILEf >a >      That _almost_ works...w    @ Well, it actually works fine for what he asked. He asked for theC argument of the LOG qualifier, not the log file-spec. OK, it adds aiA .LOG if something was specified for /LOG without a file type, but  otherwise it does what he asks.F  " OK, I guess that's nitpicking. :-]    A >      If there was no /LOG qualifier on the SUBMIT, the LOG_FILE>@ > symbol above returns "".  If the /LOG specified a logical nameE > only, e.g. SYS$SCRATCH:, LOG_FILE comes back with SYS$SCRATCH:.LOG;t, > ...only with SYS$SCRATCH fully translated. >uC >      This is an old problem that is difficult to get right in all F > cases.  Without actually testing, I think you'd need some additional- > processing long the lines of the following:t >c< > $ THIS_FILE = F$Parse(F$Environment("PROCEDURE"),,,"NAME")C > $ LOG_FULL  = F$Parse(LOG_FILE,"''THIS_FILE'.LOG;","SYS$SCRATCH")s > E > where I've specified SYS$SCRATCH because that defaults to SYS$LOGIN : > and is the default location for writing log files, IIRC.     It's SYS$LOGIN.:   So if you want the filespec:   Change your LOG_FULL to.  9 $ LOG_FULL = F$PARSE(LOG_FILE,"SYS$LOGIN:.LOG",THIS_FILE)O    B Now, if you get really picky, what if you have multiple batch jobsG running concurrently with the same (version-number-less) log file spec?iG Then it becomes non-trivial getting the correct version number for eachtF job. You'd either have to scan the output of SHOW DEVICES/FILES or useA a cool macro that someone posted once just for this very purpose.    ------------------------------  # Date: Fri, 25 Feb 2005 20:58:51 GMT # From: hoff@hp.nospam (Hoff Hoffman)w Subject: Re: Logfile namep0 Message-ID: <fkMTd.651$sY7.497@news.cpqcorp.net>  ` In article <ej9u11lup726bgah3nipv153onhs4kfn4q@4ax.com>, David Gray <police@spamcop.net> writes:  E :From within a batch job how do I find out what the value of the /LOG  :qualifier was at submit time? )  H   Ask The Wizard topics including (4568) and (9130) might be of interestG   here.  There are various examples of the f$getqui lexical there, too,tF   with the latter topic having pointers to a dozen or so other topics.    N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faqeN  --------------------------- pure personal opinion ---------------------------E         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.comp   ------------------------------    Date: 25 Feb 2005 12:27:26 -0800  From: "Barry" <dysert@gmail.com> Subject: Re: Logfile nameMC Message-ID: <1109363246.843162.311700@l41g2000cwc.googlegroups.com>I  B I don't know if you're still looking, but here's something I wrote= years ago and it always worked for the cases I needed it for:-  # $ junk=f$getqui("cancel_operation") A $ job_log_null=f$getqui("display_job","job_log_null",,"this_job"):9 $ job_name=f$getqui("display_job","job_name",,"this_job")  $sI log_specification=f$getqui("display_job","log_specification",,"this_job") 	 $ mylog:=u $ if .not. job_log_null then -<   mylog=f$parse(log_specification,job_name,"sys$login:.log")   ------------------------------  % Date: Fri, 25 Feb 2005 14:39:39 -0800e, From: Ken Fairfield <my.full.name@intel.com> Subject: Re: Logfile nameq+ Message-ID: <cvo9fb$i4n$1@news01.intel.com>   
 AEF wrote: > Ken Fairfield wrote: >  >>AEF wrote: >> >>>David Gray wrote: >>>V >>>s >>>>Hi all,e >>>> >>>nC >>>>From within a batch job how do I find out what the value of theo >  > /LOG > ! >>>>qualifier was at submit time?p >>>> >>>>e.g. >>>>@ >>>>$ submit/log=gen_logs:a_log_file_with_a_long_name.log	 X.COM >>>>F >>>>When X.COM is running I want to put the value specifed on the /LOG >>>>into a symbol. >>>> >>>>$ show symbol LogFileM< >>>>    LogFile = "gen_logs:a_log_file_with_a_long_name.log" >>>>A >>>>I was fairly sure I could do this with a lexical function butc >  > cannot >  >>>>find/remember which one. >>>>
 >>>>Cheers
 >>>>	Dave.  >>>>o >>>Try this: >>>t >>>$    SET VERIFY >>>$    LOG_FILE = > 9 > F$GETQUI("DISPLAY_JOB","LOG_SPECIFICATION",,"THIS_JOB")h >  >>>$    SHOW SYMBOL LOG_FILE >> >>     That _almost_ works...t >  eB > Well, it actually works fine for what he asked. He asked for theE > argument of the LOG qualifier, not the log file-spec. OK, it adds aUC > .LOG if something was specified for /LOG without a file type, buto! > otherwise it does what he asks.P > $ > OK, I guess that's nitpicking. :-]  :     No, you're right, I didn't read the letter of the OP's4 request, I read what I *thought* he was asking.  ;-p  A >>     If there was no /LOG qualifier on the SUBMIT, the LOG_FILE @ >>symbol above returns "".  If the /LOG specified a logical nameE >>only, e.g. SYS$SCRATCH:, LOG_FILE comes back with SYS$SCRATCH:.LOG; , >>...only with SYS$SCRATCH fully translated. >>C >>     This is an old problem that is difficult to get right in allnF >>cases.  Without actually testing, I think you'd need some additional- >>processing long the lines of the following:a >>< >>$ THIS_FILE = F$Parse(F$Environment("PROCEDURE"),,,"NAME")C >>$ LOG_FULL  = F$Parse(LOG_FILE,"''THIS_FILE'.LOG;","SYS$SCRATCH")e >>E >>where I've specified SYS$SCRATCH because that defaults to SYS$LOGIN : >>and is the default location for writing log files, IIRC. >  >  >  > It's SYS$LOGIN.T       That's two for you! :-)o   > So if you want the filespec: >  > Change your LOG_FULL to  > ; > $ LOG_FULL = F$PARSE(LOG_FILE,"SYS$LOGIN:.LOG",THIS_FILE)t  9     I prefer THIS_FILE ahead of SYS$LOGIN:.LOG, but sincei4 there's no overlap, either order will work.  And the) "SYS$LOGIN:.LOG" *is* the "default-spec".c  :     What actually amazes me is that I got it right without3 testing!  Well, except for the SYS$SCRATCH stuff...o  D > Now, if you get really picky, what if you have multiple batch jobsI > running concurrently with the same (version-number-less) log file spec?rI > Then it becomes non-trivial getting the correct version number for eacheH > job. You'd either have to scan the output of SHOW DEVICES/FILES or useC > a cool macro that someone posted once just for this very purpose.h       Precisely.   	-Kene -- n6 I don't speak for Intel, Intel doesn't speak for me...  
 Ken Fairfield7! D1C Automation VMS System Support " who:   kenneth dot h dot fairfield where: intel dot com   ------------------------------    Date: 25 Feb 2005 15:30:11 -0800$ From: "AEF" <spamsink2001@yahoo.com> Subject: Re: Logfile nameaB Message-ID: <1109374211.037579.98600@g14g2000cwa.googlegroups.com>   Barry wrote:D > I don't know if you're still looking, but here's something I wrote? > years ago and it always worked for the cases I needed it for:( >n% > $ junk=f$getqui("cancel_operation")aC > $ job_log_null=f$getqui("display_job","job_log_null",,"this_job") ; > $ job_name=f$getqui("display_job","job_name",,"this_job")- > $- >-I log_specification=f$getqui("display_job","log_specification",,"this_job")0 > $ mylog:=   > $ if .not. job_log_null then -> >   mylog=f$parse(log_specification,job_name,"sys$login:.log")    F Yikes! I forgot about the case in which the user specifies a job name.2 Yes, this is better than using the procedure name.   ------------------------------  % Date: Fri, 25 Feb 2005 20:17:24 +0100 , From: "Dr. Dweeb" <5msg0h202@sneakemail.com>< Subject: Re: London Stock Exchange slowing moving to Windows- Message-ID: <cvntkj$2dgp$1@news.cybercity.dk>s   Main, Kerry wrote: >> -----Original Message-----kC >> From: johnhreinhardt@yahoo.com [mailto:johnhreinhardt@yahoo.com] # >> Sent: February 23, 2005 10:54 PMY >> To: Info-VAX@Mvb.Saic.Com? >> Subject: Re: London Stock Exchange slowing moving to Windowss >> >> >> John Smith wrote:# >>> johnhreinhardt@yahoo.com wrote:  >>>> rob kas wrote: ( >>>>> http://www.vnunet.com/news/1161471 >>>>G >>>> "'The .Net framework really adds a lot of productivity benefits toN >> the6 >>>> developers, and we've seen it already. Instead of >> mid-size projectsD >>>> costing 1m to 2m and taking nine to 12 months, the quotes I'm >> nowE >>>> getting are hundreds of thousands of pounds, taking three to six  >>>> months to do."  >>>>C >>>> Is he really comparing the time and cost of completed projectss	 >> againt @ >>>> estimates quoted by developers????  What planet is he from? >>>r >>>eH >>> Lots of projects in the financial world start at those pricetags and >> more./ >>> I've been involved in plenty @ $5MM and up.- >>>- >>H >> I wasn't questioning the pricetag, I was having a hard time believingH >> that this guy was taking "quotes" from developers on how long and howC >> much they think a project will take at face value.  I'ts been my=6 >> experience that often they are woefully optimistic. >> >>= > And 99 times out of a 100 these cost quotes do not include:HC > - Operations costs associated with moving to new environment e.g.iE > Virus utilities, backup, batch, defrag, system monitoring and so on D > .. In the case of Windows all additional costs since prod versions# > these are not provided by the OS._G > - training of Operations staff to new technologies. And before anyone F > says "just get new staff", think about separation packages for staffC > that have been there 10-15-20 years (and in Europe these are hugeT > pkgs)AC > - cost of interface packages required to link in external vendors  > G > When I hear some Customers talk about switching platforms because appIA > development is "perceived" to be quicker (remember they have no A > mission critical record where they proved this yet), but do not D > discuss these other costs, I have to wonder "where have all the IT > Leaders of the past gone?"   they are all unemployed !!!T   > F > OTOH, personal opinion, but I have to smile sometimes at things likeE > this as it is sometimes like talking to teenagers - you try to give C > them the benefit of your experience, but they just have to try it $ > themselves before they believe it. >eB > Ah well, should make for interesting press when they do go live. >   D Indeed.  Trading on MS technology for the LSE - not a snowball IMHO.  	 Dr. DweebT   > :-)  >A	 > Regards  >  > Kerry Main > Senior Consultants > HP Services Canada > Voice: 613-592-4660a > Fax: 613-591-4477h > kerryDOTmainAThpDOTcom > (remove the DOT's and AT)e >a& > "OpenVMS has always had integrity ..  > Now, Integrity has OpenVMS .."   ------------------------------  % Date: Fri, 25 Feb 2005 14:25:29 -0500s' From: "Main, Kerry" <kerry.main@hp.com> < Subject: RE: London Stock Exchange slowing moving to WindowsR Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB53F847@tayexc19.americas.cpqcorp.net>   > -----Original Message------ > From: John Smith [mailto:a@nonymous.com]=20o" > Sent: February 25, 2005 11:30 AM > To: Info-VAX@Mvb.Saic.Com > > Subject: Re: London Stock Exchange slowing moving to Windows >=20 > Main, Kerry wrote:  
 [snip ...]    H > > OTOH, personal opinion, but I have to smile sometimes at things likeG > > this as it is sometimes like talking to teenagers - you try to giveuE > > them the benefit of your experience, but they just have to try it.& > > themselves before they believe it. > >TD > > Ah well, should make for interesting press when they do go live. >=20 >=20 > Nice attitude. >=20$ > It should have never come to this. >=20@ > HP should have been in there clawing tooth and nail showing=20 > them why theyr> > are making a big mistake switching to Windows....or is it=20 > that HP saw dollar@ > signs in selling them a truckload of Proliants using an o/s=20 > that HP makes no
 > money from?1 >=20 >=20  5 I have no idea of whether HP was involved at all. =20-  G My personal experience has been that vendors can influence a small bit,n> but in the end, typically have minimal impact on what Customer developers want to do.=20t  G In many cases, it is often what the Dev Managers personal background isiG that determines and/or at least greatly influences the direction taken.0   Regardsr  
 Kerry Main Senior Consultanto HP Services Canada Voice: 613-592-4660s Fax: 613-591-4477h kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  $ "OpenVMS has always had integrity .. Now, Integrity has OpenVMS .."   ------------------------------  % Date: Fri, 25 Feb 2005 15:24:55 -0500-# From: "John Smith" <a@nonymous.com> < Subject: Re: London Stock Exchange slowing moving to Windows, Message-ID: <Mo6dnZklMNUKFILfRVn-gg@igs.net>   Main, Kerry wrote: >> -----Original Message-----E+ >> From: John Smith [mailto:a@nonymous.com]o# >> Sent: February 25, 2005 11:30 AMc >> To: Info-VAX@Mvb.Saic.Com? >> Subject: Re: London Stock Exchange slowing moving to WindowsB >> >> Main, Kerry wrote:  >  > [snip ...] >  > H >>> OTOH, personal opinion, but I have to smile sometimes at things likeG >>> this as it is sometimes like talking to teenagers - you try to give0E >>> them the benefit of your experience, but they just have to try itE& >>> themselves before they believe it. >>>hD >>> Ah well, should make for interesting press when they do go live. >> >> >> Nice attitude.0 >>% >> It should have never come to this.1 >>> >> HP should have been in there clawing tooth and nail showing >> them why they< >> are making a big mistake switching to Windows....or is it >> that HP saw dollarL> >> signs in selling them a truckload of Proliants using an o/s >> that HP makes noo >> money from? >> >> >b3 > I have no idea of whether HP was involved at all.2 > D > My personal experience has been that vendors can influence a smallE > bit, but in the end, typically have minimal impact on what CustomerE > developers want to do. >2F > In many cases, it is often what the Dev Managers personal backgroundE > is that determines and/or at least greatly influences the directionU > taken.    , Ditto for advertising & effective marketing.F Maybe next time you are in town we can get together, down a pitcher of8 martini's and then you'll tell me what you really think.   --- OpenVMS - The classics never go out of style.    ------------------------------  % Date: Fri, 25 Feb 2005 15:18:21 -0800D3 From: "C.W.Holeman II" <cwhii_googlespam@yahoo.com> < Subject: RE: London Stock Exchange slowing moving to Windows0 Message-ID: <111vcht2uk9mh3f@corp.supernews.com>   Main, Kerry wrote:  I > My personal experience has been that vendors can influence a small bit, @ > but in the end, typically have minimal impact on what Customer > developers want to do. > I > In many cases, it is often what the Dev Managers personal background is I > that determines and/or at least greatly influences the direction taken.   H Which is why it is so important to have a strong presence in small shopsJ so that there is a larger number of Dev Managers that have lived with VMS.H So, this is true in the universities also. It is a long term project. It& requires an investment for the future.   -- C.W.Holeman II) cwh5ii@Julian5Locals.com remove the fives    ------------------------------  % Date: Fri, 25 Feb 2005 19:53:29 -05003- From: JF Mezei <jfmezei.spamnot@teksavvy.com>t< Subject: Re: London Stock Exchange slowing moving to WindowsB Message-ID: <1109378472.f435e8116246c39634aaf8109cf7c625@teranews>   "Main, Kerry" wrote:3 > I have no idea of whether HP was involved at all.     C For such a high profile account, every HP employee should have been H mobilized to save this account. HP could have pitched VMS, HP-UX, TandemH and Linux in competing bids and try to find an/all possible solutions to! retain this high profile account.   I > My personal experience has been that vendors can influence a small bit,T@ > but in the end, typically have minimal impact on what Customer > developers want to do.  D I am very surprised that you of all people would think that. I wouldG have though you would have been fully aware of the influcence IBM salesu? rep and VP have on such customers. They establish real personal H relationships with the key people in companies and make damned sure that? the relationship goes far beyond official vendor presentations.   D If HP thinks as you do, then it is no wonder they lost that account.  I > In many cases, it is often what the Dev Managers personal background iseI > that determines and/or at least greatly influences the direction taken.u  G When IBM encounters such a un-cooperative manager or even uncooperativesG techy, they wkno where to go to bypass all of this. They go to the CEO,u? CIO and discredit that uncooperative manager,s opinions (or theF discredit the person himself).  F When I pitched a disaster tolerant solution for a bank's SWIFT system,E word got to me that the CEO of the bank had been told that my projectsD wasn't possible and that I was selling something which didn't exist.E (propaganda done by IBM which didn't wan't to be seen as unable to dor  something a lowly VAX could do).  E Luckily, and by chance, I had met John Covert at a DECUS function theTD week before and he had been involved in the volume shadowing and hadH told me that what I intended to do was exactly what volume shadowing hadG been intended to do and that it would work. So when I told my superiorsdH that I had spoken to the engineer involved with that software and he hadC confirmed it would work and they had already tested such configs, I.H regains credibility. But there was never anyone from Digital who came to try to sell the solution.   G IBM was always there to influence deicions,. Digital was never there. IeH gave Digital an in to pitch their EDI solution at a time when IBM didn'tJ have a solution that worked with a VAN in canada, and DEC never showed up.  @ If HP loses London Stock Exchange, it is entirely its own fault.   ------------------------------  % Date: Fri, 25 Feb 2005 22:31:34 -0500 ' From: Dave Froble <davef@tsoft-inc.com>p< Subject: Re: London Stock Exchange slowing moving to Windows0 Message-ID: <111vqrrnrk3okce@corp.supernews.com>   JF Mezei wrote:m > "Main, Kerry" wrote: > 3 >>I have no idea of whether HP was involved at all.  >  >  > E > For such a high profile account, every HP employee should have beeneJ > mobilized to save this account. HP could have pitched VMS, HP-UX, TandemJ > and Linux in competing bids and try to find an/all possible solutions to# > retain this high profile account.r >  > I >>My personal experience has been that vendors can influence a small bit, @ >>but in the end, typically have minimal impact on what Customer >>developers want to do. >  > F > I am very surprised that you of all people would think that. I wouldI > have though you would have been fully aware of the influcence IBM saleseA > rep and VP have on such customers. They establish real personal-J > relationships with the key people in companies and make damned sure thatA > the relationship goes far beyond official vendor presentations._ > F > If HP thinks as you do, then it is no wonder they lost that account. >  > I >>In many cases, it is often what the Dev Managers personal background isaI >>that determines and/or at least greatly influences the direction taken.n >  > I > When IBM encounters such a un-cooperative manager or even uncooperativeaI > techy, they wkno where to go to bypass all of this. They go to the CEO,nA > CIO and discredit that uncooperative manager,s opinions (or then  > discredit the person himself).  F Yep!  Been there.  Seen that.  You have to expect it, and be ready to = counter it.  When you can get real info on the competition's dA limitations, that helps too, but don't be wrong or you then lose - credibility.   Dave   ------------------------------  % Date: Fri, 25 Feb 2005 22:26:05 -0500 ' From: Dave Froble <davef@tsoft-inc.com> < Subject: Re: London Stock Exchange slowing moving to Windows0 Message-ID: <111vqhjk5mk0mba@corp.supernews.com>   John Smith wrote:   K > HP should have been in there clawing tooth and nail showing them why they N > are making a big mistake switching to Windows....or is it that HP saw dollarN > signs in selling them a truckload of Proliants using an o/s that HP makes no
 > money from?.  I The real problem, if indeed HP is even involved, and has no problem with kH a truckload of Proliants, is that the customer will get a HW quote from I Dell.  Then either HP loses the business, or loses money on the business   trying to underbid Dell.  F You have to figure that if the customer is looking to cut costs, then ? they're looking to cut costs as much as possible, including HW.    ------------------------------    Date: 25 Feb 2005 13:38:54 -06004 From: kaplow_r@encompasserve.org.TRABoD (Bob Kaplow)& Subject: Re: Mfg Date of an Alpha 21003 Message-ID: <o+LRe717GGda@eisner.encompasserve.org>7  n In article <1109273886.684893.101790@o13g2000cwo.googlegroups.com>, "Shawn" <shawnm1964@sbcglobal.net> writes:H > Does anyone know when the Alpha 2100 5/250 was actually first made andG > sold to the Corporate World.  I know they are old, but management has 1 > requested that I put a year on how old ours is?e  F Take a look at the DEC serial number of the box. The format of old DECI serial number is LLYWWXXXXX where LL is the location it was made Y is the-J last digit of the year (5 for 1995) WW is the week of the year (01-52) andC XXXXX is a unique identifier within that location, year, and month.P  I Alas for Cpq and HP it's no longer so easy to tell when things were made.U  I Many demiconductors are also marked with a year/week code as well. If you J find a chip dated 9508 it was made about 120 years ago. The box itself canI only be as old as the newest chip it contains (ignoring stuff that's been  replaced since it was new).   1 	Bob Kaplow	NAR # 18L	TRA # "Impeach the TRA BoD" & 		>>> To reply, remove the TRABoD! <<<K Kaplow Klips & Baffle:	http://nira-rocketry.org/LeadingEdge/Phantom4000.pdflL     www.encompasserve.org/~kaplow_r/    www.nira-rocketry.org    www.nar.org  K I'm not afraid of terrorists.  I am terrorized by airport security. _george    ------------------------------  % Date: Fri, 25 Feb 2005 16:34:12 -0500A( From: Bill Todd <billtodd@metrocast.net>A Subject: Re: OpenVMS Seminar in Toronto (2005-02-24) a few pointsh= Message-ID: <avWdnRTVc4dJBILfRVn-1A@metrocastcablevision.com>   G I passed up on responding to this first time around, but since Rob has AB seen fit to parrot it over at RealWorldTech I'll just recycle the  response I gave there.   Neil Rieck wrote:l   ...a  K  >OpenVMS business is still growing at ~12% per year since the HP merger inA  >2002.   C Wow! At that rate, a few more years and it'll be back where it was  G before the Alphacide (which cut overall VMS system revenues, including II those from on-going service contracts, in half, and thus cut sales by an   even greater percentage).E  E Having been restored to that level, it'll then be about where it was RG when Mike Capellas came ever-so-close to scheduling it for mothballing _ back in mid-1999.    ...T    >OpenVMS is #1 in health care8  >OpenVMS is 90% in chip manufacturing [guesses anyone?](  >OpenVMS is 50% in mobile phone billing(  >OpenVMS is 50% of all SMS transactions1  >OpenVMS is 66% of all electronic bank transfers40  >OpenVMS runs 5 out of the top 10 stock markets  = Wow again! Now, these wouldn't be like the perpetual 411,000 2F installed-system counts that cHumPaq has been claiming for going on 5 E years straight now, would they? Funny how just failing to update old eF data can present an image of continued health in what appears well on  the road to becoming a corpse.  I (That last may require some expansion:  the context of the discussion at eI RWT was in part the suggestion by Charlie ('Groo') that his sources told  @ him that VMS staffing was being markedly reduced, moving toward F something looking an awful lot like maintenance levels.  Since that's H the same thing others have told me, it's not very difficult to believe.)   - bill   ------------------------------  % Date: Fri, 25 Feb 2005 20:16:32 -0500h- From: JF Mezei <jfmezei.spamnot@teksavvy.com>aA Subject: Re: OpenVMS Seminar in Toronto (2005-02-24) a few pointsiB Message-ID: <1109379855.97a4f2b49f0d83d805ddfa8693b8fff0@teranews>   Neil Rieck wrote:cK > OpenVMS business is still growing at ~12% per year since the HP merger ine) > 2002. (Seems to be a big hit in Europe)-1 > OpenVMS is 66% of all electronic bank transferss    D This statistic is not credible, by far, and devalues the validity of other statistics.e  0 > OpenVMS runs 5 out of the top 10 stock markets  A Not for much longer if HP is helping stock markets such as LSE to  migrate to Windows.r  H > OpenVMS Itanium licensing will be simpler and cheaper. The three major > categories are:d > FOE - Foundation > EOE - Enterprise > MCOE - Mission Critical-/ > (languages will still be licensed separately)0  F Is there any "workstation" class licencing ? (Hoff has been so adamentE that VMS would remain available in workstation-configured servers, soe one must ask about licencing).   ------------------------------  % Date: Fri, 25 Feb 2005 22:42:17 -0500r' From: Dave Froble <davef@tsoft-inc.com> A Subject: Re: OpenVMS Seminar in Toronto (2005-02-24) a few points 0 Message-ID: <111vrg1aiknd7d9@corp.supernews.com>   Neil Rieck wrote:u > Folks, > H > Yesterday (2005-02-24) I attended the OpenVMS "Technical Outlook 2005"G > seminar in Toronto. The slides should be posted at the following URL  
 > sometime > soon: ! > http://www.encompasscanada.com/ 3 > but here are "a few" highlights from my note pad.t >  > * * * * *  > K > OpenVMS business is still growing at ~12% per year since the HP merger ini) > 2002. (Seems to be a big hit in Europe)i  G I'm not sure what this statistic means.  If just looking at the number pH of new systems sold, then great, more systems are being sold.  Or is it A dollars of new system business.  If based upon dollars, then the oI quantity may still be low, and that isn't really growth of usage of VMS. gC   If most of the sales are to existing accounts, then retention of  G existing accounts is also a big question.  Suppose that for every sale aF of a new system, 5 existing customers are moving off VMS.  There goes  the after sales dollars.  I I don't think one liners say much.  In an environment desperate for good o news they can be misleading.  1 I think the descending order goes something like:t   1) Lies  2) Damn lies
 3) Statisticsg  G Tell me that the number of customers purchasing software and haardware -3 support is rising, and I'll get a bit more excited.e   Dave   ------------------------------  % Date: Fri, 25 Feb 2005 20:25:34 -0600g2 From: David J Dachtera <djesys.nospam@comcast.net>( Subject: Re: Ouch! a *MAJOR* bug in TPU.+ Message-ID: <421FDE1D.9D3ECBF3@comcast.net>o   Bart Zorn wrote: > g > David J Dachtera <djesys.nospam@comcast.net> wrote in message news:<421E99A1.9ED7F273@comcast.net>...t >  >    [ S n i p . . . ] > J > > First, I'd consider the NFS store to be archival only. Before editing,J > > make a local copy, and don't replace the source until you're satisfied > > with the changes.r > 	 > Agreed.  > H > > Second, lines longer than the width of the screen are unsuitable forH > > common practice use with a text editor. If the application cannot beH > > "fixed" to use a more reasonable configuration file format, considerJ > > developing a configuration editor (and make it as "portable", whatever, > > that means to you, as your application). > 	 > Agreed.r > 5 > > By the way: TPU can cause inadvertent corruption:/ > >e > > o Edit any file.L > > o Move the cursor to the end of any line where the next line is shorter.. > > o Cursor down once, and hit the space bar. > >.H > > You just corrupted your file. The cursor line is now one byte longerG > > than the line above it. TPU/EVE has space-padded the line up to theIJ > > cursor position before you added a space, and then added the space you > > just typed.  > G > I would not call this corruption. It is documented behaviour and youriE > file is certainly not fubar. You can use the EVE command SET CURSOR ' > BOUND to prevent this from happening.I >  > > Fun stuff, huh?  > >lL > > TPU/EVE has its uses, don't get me wrong. So has EDT, vi, eMACS and evenH > > EDLIN (if you still have any MS-DOS or W/9X system running). Just be7 > > aware of what the program can do you and your text.u >  > I agree mostly, but vi ?  B WHen I was introduced to it circa. 1986, it was a screen-oritentedH editor on a UN*X system that otherwise lacked one. I made it as EDT-like as I could, within my limits.t   -- , David J Dachtera dba DJE Systems. http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:u" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/a  " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/e   Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  # Date: Sat, 26 Feb 2005 02:27:49 GMTc  From: John Santos <john@egh.com>( Subject: Re: Ouch! a *MAJOR* bug in TPU.+ Message-ID: <F8RTd.13042$QQ3.5578@trnddc02>-   Bob Koehler wrote:j > In article <cvl6rk$jt5$1@hercules.btinternet.com>, "Richard Maher" <maher_rj@hotspamnotmail.com> writes: >  >>Hi,. >>K >>I didn't see the original problem, but along the lines of "*MAJOR* bug in # >>TPU" how's this for a feature : -i >>N >>1) You edit some code with /READ 'cos the source in your reference directory6 >>and you haven't checked it out for modification yet. >  > E >    /READ marks the buffer both NO MODIFY and NO WRITE.  If you onlyeI >    change NO MODIFY you get what you asked for, a buffer which TPU willnJ >    let you modify, but will not automatically write.  I have buttons to I >    toggle each of these separately in my EVE interface for exactly this  >    reason. > K >    Yes, there are many legitimate times when I want buffers MODIFY and NOeG >    WRITE.  But if you don't then write a little TPU routine to toggle  >    both at the same time.n  H I'm not a TPU user (usually), but I think the issue here is not that TPUD doesn't automatically clear the "NO WRITE" setting when you manuallyG change the "NO MODIFY" setting and modify the buffer, but that TPU lets-D you exit when you have modified buffers and no where to put them.  IE think the OP would be happy if it said "You've got a modified buffer,DB do you really want to exit", and you could say "No", then save theG modified file to a new version of the existing file or whatever.  As ite> is, it just exits and all your hard work is in the bit-bucket.  > Maybe there is a bit of magic that can be done in the OP's TPUD initialization file that would do exactly this?  (Force confirmation; before exiting with modified, unwritable buffers, that is.)i  @ TECO says "?NFO    No file for output" in similar circumstances.   -- s John Santosa Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Fri, 25 Feb 2005 22:05:10 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> ( Subject: Re: Ouch! a *MAJOR* bug in TPU.B Message-ID: <1109386364.5bd3f915e1b5980c87c2c36df3cbed1b@teranews>   John Santos wrote:@ > Maybe there is a bit of magic that can be done in the OP's TPUF > initialization file that would do exactly this?  (Force confirmation= > before exiting with modified, unwritable buffers, that is.)m  C No. TPU/EVE needs to be modified so everyone benefits from having a[G confirmation of unsaved changes when you have  amodifiable buffer beingXD asked to wrote to a unwritable file to advise that this operation isD impossible instead of just quitting from TPU without any warning and losing the changes.    ------------------------------  % Date: Fri, 25 Feb 2005 21:03:41 -0600j2 From: David J Dachtera <djesys.nospam@comcast.net>( Subject: Re: Ouch! a *MAJOR* bug in TPU.+ Message-ID: <421FE70D.BF587959@comcast.net>i   Bob Koehler wrote: > b > In article <421E99A1.9ED7F273@comcast.net>, David J Dachtera <djesys.nospam@comcast.net> writes: > > H > > You just corrupted your file. The cursor line is now one byte longerG > > than the line above it. TPU/EVE has space-padded the line up to thetJ > > cursor position before you added a space, and then added the space you > > just typed.  > D >    You just edited your file, telling TPU to put a space where the2 >    cursor is.  That's hardly the editor's fault.  E Well, actually, yes it is. See, the editor failed to first repositionDF the cursor to the end of the line, which (IMO) it should have done. ItG then added data I did not intend before adding the data I *DID* intend.o  3 I believe that is *THE* definition of "corruption".s  H >    Yes, I know oher editors can and do use the space bar as a movement? >    command.  Tradiationaly DEC's editors didn't and TPU is no  >    exception.i     -- n David J Dachtera dba DJE Systemsf http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:s" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/"  " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/G   Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Sat, 26 Feb 2005 03:30:57 -0000i6 From: "Alex Daniels" <AlexNoSpamDaniels@themail.co.uk> Subject: Re: Point Secured6 Message-ID: <421fed74$0$32615$db0fefd9@news.zen.co.uk>  , <VAXman- @SendSpamHere.ORG> wrote in message* news:00A3FEC6.E08F52DD@SendSpamHere.ORG...K > In article <EdJTd.21865$bp3.8833@fe03.lga>, "Dan Moore" <dmoore@sosu.edu>i writes:R? > Is there another product that offers all of these features incL > >one package? I'm also interested to hear about system performance impact, > >which I'm told is minimal.U >t > SecurityGuard! >P > http://www.provn.com/. >" >I > -- O2 > VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)COM > 6 >   "Well my son, life is like a beanstalk, isn't it?"  @ I use the Intruder Alert and AUDIT parts of it, would recommend.   Alex   ------------------------------  % Date: Fri, 25 Feb 2005 14:44:56 -0500u( From: Bill Todd <billtodd@metrocast.net> Subject: Re: Sayonara Tukwilla= Message-ID: <oZadnWRsyvemHYLfRVn-tw@metrocastcablevision.com>5   Rob Young wrote:   ...m  " > 	Superdome is long in the tooth.  H Superdome was getting long in the tooth when McKinley released nearly 3 
 years ago.      Surely something new in the? > 	wings, but of course not much chatter or it freezes existingw	 > 	sales.o  I Surely?  Well, if you say so:  Carly was such a visionary that of course  A she'd have given priority to such an effort starting right after eC Superdome originally shipped in Y2K (since it had the same scaling  ( failings that graced Alpha's GS series).  E Then again, perhaps the same server people who helped trash EV8 were <C whispering in her ear that Itanic would cure all ills and that the eG server architecture (with its sx1 face-lift) was just fine, thank you. -I They certainly convinced Robison of that at Compaq, so Carly should have ) been a push-over.E   - bill   ------------------------------  % Date: Sat, 26 Feb 2005 03:50:19 +08002 From: prep@prep.synonet.comd Subject: Re: Sayonara Tukwilla- Message-ID: <87u0o02zd0.fsf@prep.synonet.com>1  ) Dave Froble <davef@tsoft-inc.com> writes:    > JF Mezei wrote:l >> icerq4a@spray.se wrote: >>G >>>The new Tukwila will be a 4-core Montvale product with the EV8 teams 8 >>>system infrastructure, and that is what Itanium need.F >> No, this article says that they dumped the idea of 4 cores and will >> keepj3 >> it at 2 core, and just try to improve the cores.i >i- > So let me see if I've got the rumors right.  >oE > DEC engineers are designing a chip, which will have 8 cores, and on, > chip SMP.$ >iG > HP engineers don't want the CPU design of the DEC engineers, they get9F > the project taken away from the DEC engineers, they'll keep the SMP,( > but re-do the CPU to their own liking. >1 > Now they're down to 2 cores. >a" > Now they're all Intel engineers. >>
 > That right?t  I You missed the bit about the Alpha team having taped out 1st pass siliconIC and planing the second pass. Other than that, seems pretty spot on.t   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Sat, 26 Feb 2005 03:07:05 -0000e6 From: "Alex Daniels" <AlexNoSpamDaniels@themail.co.uk> Subject: Re: Sayonara Tukwilla6 Message-ID: <421fe7db$0$32619$db0fefd9@news.zen.co.uk>  5 "Bill Todd" <billtodd@metrocast.net> wrote in messagem7 news:oZadnWRsyvemHYLfRVn-tw@metrocastcablevision.com...w1 > priority to such an effort starting right afterOD > Superdome originally shipped in Y2K (since it had the same scaling* > failings that graced Alpha's GS series).    Do you just mean Wildfires here?  ( Or are you including the GS1280 in this?   Alex   ------------------------------  % Date: Fri, 25 Feb 2005 22:10:46 -0500 ( From: Bill Todd <billtodd@metrocast.net> Subject: Re: Sayonara Tukwilla= Message-ID: <i_GdnZ_i2eYqdYLfRVn-sA@metrocastcablevision.com>f   Alex Daniels wrote:o7 > "Bill Todd" <billtodd@metrocast.net> wrote in messagek9 > news:oZadnWRsyvemHYLfRVn-tw@metrocastcablevision.com...t > 1 >>priority to such an effort starting right afteriD >>Superdome originally shipped in Y2K (since it had the same scaling* >>failings that graced Alpha's GS series). >  > " > Do you just mean Wildfires here?  G Yes - in a moment of brain-fade I forgot that the designation had been $ carried forward.   > * > Or are you including the GS1280 in this?  	 Hell, no.S   - bill   ------------------------------  % Date: Fri, 25 Feb 2005 22:51:21 -0500n' From: Dave Froble <davef@tsoft-inc.com>S Subject: Re: Sayonara Tukwilla0 Message-ID: <111vs1065q9tia2@corp.supernews.com>   Bill Todd wrote: > Rob Young wrote: >  > ...E > & >>     Superdome is long in the tooth. >  > J > Superdome was getting long in the tooth when McKinley released nearly 3  > years ago. >  >   Surely something new in theh > C >>     wings, but of course not much chatter or it freezes existingw
 >>     sales.! >  > K > Surely?  Well, if you say so:  Carly was such a visionary that of course iC > she'd have given priority to such an effort starting right after (E > Superdome originally shipped in Y2K (since it had the same scaling t* > failings that graced Alpha's GS series). > G > Then again, perhaps the same server people who helped trash EV8 were o7 > whispering in her ear that Itanic would cure all illso  H I thought that these same server people who, after trashing Alpha, were F themselves trashed as soon as HP took over?  It's only poetic justice.   > and that the tI > server architecture (with its sx1 face-lift) was just fine, thank you.  K > They certainly convinced Robison of that at Compaq, so Carly should have c > been a push-over.   F What's really disgusting is that Robinson wasn't trashed when HP took C over.  It was my observation that it wasn't so much Robinson being gH convinced, as the server people rubber stamping what Robinson wanted in C order to keep their jobs.  That made their getting trashed so much uH sweeter.  Who Robinson was shilling for is anyone's guess.  The desires  of the BOD?  Winkler?  Curly?-  < Then again, maybe I'm totally out of touch with any reality.   Dave   ------------------------------  % Date: Fri, 25 Feb 2005 14:19:01 -0600$# From: "Dan Moore" <dmoore@sosu.edu>$ Subject: Re: ssl & osu) Message-ID: <cSLTd.1362$6d6.971@fe07.lga>s   Chuck,  J     STUNNEL may work. It's on the OpenVMS freeware cd. Basically, stunnel L will listen for ssl (https:) connections on a given tcp port, then redirect F the connection to another ip/port. So, the https: connection (running K stunnel) on the secure ip would be redirect locally (and transparently) to e  the web server ip (running osu).   Example:5     1. PC connects to https://192.168.58.1 (port 443)yJ     2. 192.168.58.1:443 redirects to 192.168.58.2:80 (via stunnel locally)  0 (modified from the stunnel readme_vms.txt file):G A. On the server, the server command (in STUNNEL_STARTUP_SERVER.COM) iscI     $ stunnel_command = "STUNNEL -d 443 -r 192.168.58.2:80 -p stunnel.pem   K This may be the long way, I'm not familiar with OSU, which probably has an  1 easy way to configure SSL vs. non-SSL separately.    Dan     9 "Chuck Aaron" <caaron@ceris.purdue.edu> wrote in message n/ news:cvnah0$18i$1@mailhub227.itcs.purdue.edu...i > Group, >dJ > I am running version 3.10a osu web server with openvms SSL on VMS 7.3-2. > Our users access our server G > based on two different ip addresses. Do you know if there is a way toy" > make users who are accessing oneH > of the ip address applications on the web to force them to ssl (https)& > and leave the other application http< > or will I have to make it all https if I chose that route? >i	 > Thanks,r >t
 > Chuck Aarong >h >    ------------------------------  # Date: Sat, 26 Feb 2005 00:17:15 GMToL From: winston@SSRL.SLAC.STANFORD.EDU (Alan Winston - SSRL Central Computing) Subject: Re: ssl & osu6 Message-ID: <00A3FEDD.EF427C7C@SSRL.SLAC.STANFORD.EDU>  c In article <cvnah0$18i$1@mailhub227.itcs.purdue.edu>, Chuck Aaron <caaron@ceris.purdue.edu> writes:n- >This is a multi-part message in MIME format.,' >--------------090507020305070600080107o: >Content-Type: text/plain; charset=us-ascii; format=flowed  >Content-Transfer-Encoding: 7bit >d >Group,S > J >I am running version 3.10a osu web server with openvms SSL on VMS 7.3-2.    >Our users access our serverG >based on two different ip addresses. Do you know if there is a way to r! >make users who are accessing one H >of the ip address applications on the web to force them to ssl (https) % >and leave the other application httpl; >or will I have to make it all https if I chose that route?e   So:"   address-1: HTTPS access only   address-2: http access  - and these are different numeric IP addresses.   F Do they have different directory trees?  If so, life is fairly easy.      H If you just want non-SSL access to fail for the stuff on address-1, you J can set up a PROTECTION statement for the stuff in the address-1 directory3 tree.  (You may as well put it in http_paths.conf.)   5 hostprot /address1-root/*  www_system:https_only.prot   / and www_system:https_only.prot would look like:e     maskgroup @*.*.*.*:443    O (This gives access to anybody anywhere so long as they use the https port.  You N can adjust the ip address octet parts to restrict it to your local subnet, and? you can make other additive restrictions like requiring login.)3    N If you want non-SSL access to be redirected to the same address but using SSL,7 that's easy if you're already using localddress blocks:I   You'll want something like   localaddress address-1-name A  redirect http://address1-name.tld/*  https://address1-name.tld/*m localaddress address-2 etc.  I (I don't have time now to check on whether you can use a wildcard for theEL nodename portion of the fully qualified URL; if you can't, you might need to3 have redirects in your localaddress block for both:    http://address1-nodename/* and 2   http://fully-qualified-domain-name-of-address-1/D because the redirect template is going to do a plain textual match.)  F Anyway, you can certainly do what you're asking about without forcing  address2 to be HTTPS only.   Shameless plug:i  K (This stuff is discussed in my Webservers-on-VMS book, which covers Apache,e. OSU, and WASD.  Chapters 5, 7, and 9, I think.  : http://www.hp.com/hpbooks/digital_press/dp_1555582648.html  K Feel free to buy it, but I hope I've given you enough to get you going.  If $ you need more guidance, please ask.)     -- Alanl   ------------------------------    Date: 25 Feb 2005 13:01:04 -0800 From: JimStrehlow@data911.coms; Subject: Re: WRQreflection 11: relocate settings directory?iB Message-ID: <1109365264.765611.87450@l41g2000cwc.googlegroups.com>   David J Dachtera wrote:1 > jqq@myrealbox.com wrote:% ... I've never heard of Reflection 112  D We use WRQ Reflection X Single User English Version 11.0 in order toF remotely use an Xwindows graphical user interface for Oracle UniversalF Installer of Oracle 9 for OpenVMS. It can be slow; but certainly beatsF having to drive to a remote location for enhancements after an initial installation of a database.n  @ Jim Strehlow, OpenVMS Systems Manager, Data911, Alameda, CA, USA   ------------------------------  % Date: Fri, 25 Feb 2005 21:12:47 -0600 2 From: David J Dachtera <djesys.nospam@comcast.net>; Subject: Re: WRQreflection 11: relocate settings directory?-+ Message-ID: <421FE92E.E4331A96@comcast.net>5   "Stuart, Ed" wrote:m > N > In the About section of the product Reflection for UNIX and Digital is named > Reflection 11.  
 What version?g  B The one I have at work shows V8.0.6, and the one I have here shows V6.00.  ' You're looking at Help -> About, right?F   --   David J Dachtera dba DJE Systems@ http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/o  " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/n   Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Fri, 25 Feb 2005 21:04:56 -0600e2 From: David J Dachtera <djesys.nospam@comcast.net>; Subject: Re: WRQreflection 11: relocate settings directory? + Message-ID: <421FE757.68CC8F33@comcast.net>>   JimStrehlow@data911.com wrote: >  > David J Dachtera wrote:t > > jqq@myrealbox.com wrote:' > ... I've never heard of Reflection 11  > F > We use WRQ Reflection X Single User English Version 11.0 in order toH > remotely use an Xwindows graphical user interface for Oracle UniversalH > Installer of Oracle 9 for OpenVMS. It can be slow; but certainly beatsH > having to drive to a remote location for enhancements after an initial > installation of a database.t  D Reflection/X V11 is rather different from Reflection for Dinital and Unix Vwhatever.e   -- o David J Dachtera dba DJE Systemss http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page:t" http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------   End of INFO-VAX 2005.113 ************************